Skip to main content
This is a legacy endpoint. Use /api/embed for new integrations.
Generate embeddings from a model for a single text prompt. This endpoint returns embeddings as float64 values.

​
Request

​
Endpoint

POST /api/embeddings

​
Request Body

​
model
string
required
Model name to use for embedding generation (e.g., nomic-embed-text)
​
prompt
string
required
Text prompt to generate embeddings for
​
keep_alive
string | number
default:"5m"
Duration to keep model loaded in memory (e.g., "5m", "1h", -1 for indefinite)
​
options
object
Model-specific options for embedding generation

​
Response

​
Response Fields

​
embedding
array
Array of float64 numbers representing the embedding vector

​
Examples

curl http://localhost:11434/api/embeddings -d '{
  "model": "nomic-embed-text",
  "prompt": "Hello, world!"
}'

​
Example Response

{
  "embedding": [
    0.5670403838157654,
    0.009260174818336964,
    -0.4829583466053009,
    0.11173559725284576,
    -0.8732880353927612,
    ...
  ]
}

​
Error Responses

​
error
string
Description of the error

​
Common Errors

  • 400 Bad Request: Model name or prompt is required
  • 404 Not Found: Model not found
  • 500 Internal Server Error: Embedding generation error
Migration Guide: To migrate from /api/embeddings to /api/embed:
  • Change prompt to input
  • Embeddings are returned in a nested array: embeddings[0] instead of embedding
  • Float32 precision instead of Float64
  • Support for batch processing with array input